mogsa (version 1.6.4)

bootMbpca: Bootstrap mbpca to estimate the coherence of different data sets

Description

Bootstrap mbpca to estimate the coherence of different data sets and estimate the number of components should be included in an analysis.

Usage

bootMbpca(moa, mc.cores = 1, B = 100, replace = TRUE,  resample = c("sample", "gene", "total"), log = "y", ncomp = NULL, method = NULL,  maxiter = 1000, svd.solver = c("svd", "fast.svd", "propack"), plot = TRUE)

Arguments

moa

An object of moa returned by mbpca.

mc.cores

Integer; number of cores used in bootstrap. This value is passed to function mclapply

B

Integer; number of bootstrap

replace

Logical; sampling with or without replacement

resample

Could be one of "sample", "gene" or "total". "sample" and "gene" means sample-wise and variable-wise resampling, repectively. "total" means total resampling.

log

Could be "x", "y" or "xy" for plot log axis

ncomp

Passed to function mbpca. In most of cases, user do not need to specify this argument because it could be inferred from moa.

method

Passed to function mbpca.In most of cases, user do not need to specify this argument because it could be inferred from moa.

maxiter

Passed to function mbpca.In most of cases, user do not need to specify this argument because it could be inferred from moa.

svd.solver

Passed to function mbpca.In most of cases, user do not need to specify this argument because it could be inferred from moa.

plot

Logical; whether the result should be plotted.

Value

It returns a matrix, columns are eigenvalues for different components. Each rows is a bootstramp sample.

Details

update details.

Examples

# see examples in \code{\link{mbpca}}